Heart rate variability and drawing impairment in hypoxemic COPD

Brain and Cognition 70 (2009) 163–170

Contents lists available at ScienceDirect

Brain and Cognition journal homepage: www.elsevier.com/locate/b&c

Heart rate variability and drawing impairment in hypoxemic COPD Raffaele Antonelli Incalzi a,e, Andrea Corsonello b,*, Luigi Trojano c, Claudio Pedone a, Domenico Acanfora d, Aldo Spada e, Gianni D’Addio d, Roberto Maestri f, Franco Rengo d,g, Giuseppe Rengo d a

Chair of Geriatric Medicine, University Campus Bio-Medico, Rome, Italy Italian National Research Center on Aging (INRCA), Viale della Resistenza Pal. Alfa Scala H, I-87036 Rende (CS), Italy c Department of Psychology, Second University of Naples, Caserta, Italy d Salvatore Maugeri Foundation, Institute of Care and Scientific Research, Rehabilitation Institute of Telese, Benevento, Italy e San Raffaele Foundation-Cittadella della Carità, Taranto, Italy f Salvatore Maugeri Foundation, Institute of Care and Scientific Research, Institute of Montescano, Italy g Chair of Geriatric Medicine, ‘‘Federico II” University School of Medicine, Naples, Italy b

a r t i c l e

i n f o

Article history: Accepted 28 January 2009 Available online 3 March 2009 Keywords: Cognitive impairment Chronic obstructive pulmonary disease Heart rate variability Neuropsychologic assessment Elderly

a b s t r a c t We studied 54 patients with hypoxemic chronic obstructive pulmonary disease (COPD). The Mini Mental State Examination and the Mental Deterioration Battery were used for neuropsychological assessment. Heart rate variability (HRV) was assessed based on 24-h Holter ECG recording. Mann–Whitney test was used to compare HRV parameters of patients performing normally or abnormally on individual neuropsychological tasks. Spearman’s rho was used to investigate the correlations between HRV parameters and neuropsychological scores, indexes of health status or COPD severity. Patients with defective performance at copying drawings with landmarks (CDL) test (N = 23) had lower very low frequency (VLF) power with respect to patients with normal performance (N = 31) (24 h: median 213; interquartile range 120–282 vs. 309; 188–431 ms2, p = 0.043; daytime: 202; 111–292 vs. 342; 194–397 ms2, p = 0.039). The CDL score correlated with the VLF power (24 h: q = 0.27, p = 0.049; daytime: q = 0.30, p = 0.028), and the normalized low frequency/high frequency (LF/HF) ratio (24 h: q = 0.27, p = 0.05; daytime: q = 0.33, p = 0.015). Sympathetic modulation decreased for increasing severity of COPD. In conclusion, drawing impairment correlates with depressed sympathetic modulation in patients with COPD, and both might be indexes of COPD severity. Ó 2009 Elsevier Inc. All rights reserved.

1. Introduction Cognitive impairment is a common finding in hypoxemic chronic obstructive pulmonary disease (COPD) and can affect a large proportion of patients. COPD-related cognitive dysfunction presents with impaired constructional abilities, verbal learning, verbal attainment and visuospatial intelligence (Antonelli Incalzi et al., 1993). This pattern of cognitive impairment might be correlated to dysfunction of anterior frontal areas, as suggested by a single photon emission computed tomography (SPECT) study showing decreased brain perfusion in hypoxemic COPD patients, particularly in anterior cortical and subcortical regions (Antonelli Incalzi et al., 1993). More recently, a further SPECT study confirmed findings of significant reduction of cerebral perfusion particularly in anterior frontal and parietal areas of the left hemisphere in a small sample of hypoxemic COPD patients (Ortapamuk and Naldoken, 2006). In both neurofunctional imaging studies brain hypoperfu-

* Corresponding author. Fax: +39 0984461872. E-mail address: [email protected] (A. Corsonello). 0278-2626/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.bandc.2009.01.010

sion was significantly correlated with worse performance on several neuropsychological tasks. Detecting cognitive impairment might have important clinical and health care implications in COPD patients, since cognitive dysfunction is associated with poor compliance with the therapy and reduced functional autonomy in several basic and instrumental activities of daily living (IADL) (Allen, Jain, Ragab, & Malik, 2003; Antonelli Incalzi et al., 1997). Most importantly cognitive dysfunction may have strong prognostic implications. Indeed, a defective performance on a drawing task, the Haecan copy to landmark test, has been found to be associated with increased mortality in hypoxemic COPD patients, while performances on other cognitive tests were not found to be associated with increased risk of death (Antonelli Incalzi et al., 2006). These findings may parallel those from studies on Alzheimer’s patients (Claus, Walstra, Bossuyt, Teunisse, & Van Gool, 1999) and on nondemented elderly (Royall, Chiodo, Mouton, & Polk, 2007) in whom drawing impairment has been reported to predict mortality independently from classical prognostic indicators. However, such studies do not clarify the mechanisms by which performance on drawing tasks can be linked to higher mortality. In COPD the

164

R. Antonelli Incalzi et al. / Brain and Cognition 70 (2009) 163–170

drawing impairment has been considered related to frontal executive dysfunction and an index of widespread involvement of brain circuits in the disease, with consequent negative prognostic value (Antonelli Incalzi et al., 2006; Claus et al., 1999), but a more specific hypothesis would link drawing impairment with a right insular dysfunction, that in its turn would induce defective cortical control of autonomic function (Royall et al., 2007). The postulated neuroautonomic dysfunction would cause a loss of sympathetic modulation: unopposed parasympathetic tone might be responsible for bradyarrhytmias and loss of balance with falls and fractures (Royall, Gao, et al., 2006). This hypothesis seems of interest also to explain the high incidence of sudden death, orthostatic hypotension, falls and fractures in Alzheimer’s patients (Kenny, Kalaria, & Ballard, 2002). Indeed, neuroautonomic dysfunction is frequent in all type of dementia, but is especially evident in patients with Alzheimer’s disease (Allan et al., 2007), and worsens with its progression (Zulli et al., 2005). In mild hypoxemic COPD patients a previous study demonstrated abnormal neuroautonomic functions and a depressed heart rate variability (HRV) response to sympathetic and vagal stimuli (Volterrani et al., 1994). Several regions of the brain, and particularly right anterior cingulate cortex and right insula, are involved in control of neuroautonomic functions (Critchley, Elliot, Mathias, & Dolan, 2000), and patients with bilateral focal lesions involving medial prefrontal areas may show abnormalities in autonomic cardiovascular responses with blunted autonomic arousal to mental stress (Critchley et al., 2003). However, some studies on normal subjects have shown that even cortical areas involved in allocation of attentional and cognitive resources, i.e. dorsolateral prefrontal cortex, may modulate HRV responses in a challenging environment (for a review, see Thayer & Brosschot, 2005). Therefore, both frontal executive and right insular dysfunction might ultimately lead to autonomic imbalance. On this basis, it is possible to hypothesize that the functional impairment of anterior frontal areas documented by SPECT studies in hypoxemic COPD patients is correlated with abnormal HRV patterns, and that such HRV abnormalities are most prominent in COPD patients showing impaired performances on the copying with landmark test. A specific pattern of correlation between drawing impairment and HRV measures (i.e. impaired neuroautonomic control) would suggest that simple neuropsychological tools may be considered as an index of the underlying neurobiological alterations, likely related to frontal dysfunction, particularly of the right hemisphere, in patients with COPD ( Royall, Gao, et al., 2006; Royall, 2006; Royall et al., 2007). In the present paper, we used data from a study on the correlation between COPD severity and HRV (Antonelli Incalzi et al., 2006) to verify whether drawing ability is associated with impaired patterns of HRV in hypoxemic COPD. The study of HRV was performed by means of the computerized analysis of variability in beat-to-beat (RR) intervals during a routine electrocardiogram recording in the 24 h. Such analysis provides specific measures of neuroautonomic functions, assessing both sympathetic and parasympathetic tone. Actually, different ranges of power spectral density of RR intervals reflect functioning of selective districts of the neuroautonomic system: the efferent vagal activity is best estimated by the high frequency (HF) component of HRV (Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, 1996), whereas the very low frequency (VLF) component reflects neuroendocrine influences mediated mainly by the sympathetic system (Takabatake et al., 2001; Vinik & Ziegler, 2007). Other measures of HRV (total power, TP; low frequency, LF; LF/HF ratio) are considered to reflect the sympatho/vagal balance (for further details, see below).

2. Methods In the context of a study on cognitive training in hypoxemic COPD (Antonelli Incalzi et al., 2008), an analysis of HRV was performed to assess the relationship between autonomic dysfunction and COPD severity. In the present study, we analysed HRV data to investigate the correlation between drawing ability and HRV. Details on the study population and the design of the parent study, are available elsewhere (Antonelli Incalzi et al., 2006; Antonelli Incalzi et al., 2008). Briefly, we consecutively enrolled 149 COPD patients who had undergone a 40–60 day period of in-hospital rehabilitation following an acute exacerbation of COPD. At the time of data collection participants were in stable conditions and their oxygen partial pressure (PaO2), carbon dioxyde partial pressure (PaCO2) and personal independence were comparable to those reported before the exacerbation in the ambulatory records. The diagnosis of COPD was made according to the American Thoracic Society standards: post-bronchodilator forced expiratory volume in one second (FEV1)/forced vital capacity (FVC) ratio was measured by spirometry and values less or equal to 0.7 were considered to confirm the presence of an airflow limitation that was not fully reversible (American Thoracic Society, 1995). Only hypoxemic patients (i.e. those with PaO2 6 56 mm Hg either at rest or at the end of the 6 min walking test) were selected because cognitive impairment is highly prevalent and clinically important in hypoxemic COPD patients (Antonelli Incalzi et al., 1993). Probable Alzheimer’s disease (McKhann et al., 1984), vascular dementia (Rosen, Terry, Fuld, Katzman, & Peck, 1980), or other diseases known to affect cognitive status (Consensus Conference, 1987) were criteria of exclusion from the study. Obstructive Sleep Apnea was excluded on the basis of a structured interview, but not of a sleep study. Fifty-four out of 149 patients were randomly selected to underwent HRV examination by simple randomization procedure at the Coordinating Center. Although this was a convenience sample, a comparable sample size had been proved adequate to disclose a correlation between HRV and cognitive performance in patients with Alzheimer’s disease or mild cognitive impairment (Zulli et al., 2005). The study protocol was approved by the Ethical Committee of Salvatore Maugeri Foundation, Institute of Care and Scientific Research, Rehabilitation Institute of Telese, Benevento, Italy. Patients gave an informed, written consent to be enrolled in the study. 2.1. Clinical and neuropsychological assessment Clinical assessment has been extensively described elsewhere, and included spirometry, arterial blood gases measurement (Radiometer ABL 500, Diamond Diagnostics, Holliston, MA, USA), 6 min walking test, Borg’s scale of dyspnea, Barthel Index, and instrumental activities of daily living (IADL) (Antonelli Incalzi et al., 2006). Comorbid diseases were coded according to the International Classification of Diseases 9th Revision – Clinical Modification (ICD9-CM) (PHS-HCF, 1980). All patients underwent a complete neuropsychological assessment including the Mini Mental State Examination (MMSE) (Folstein, Folstein, & McHigh, 1975), and the Mental Deterioration Battery (MDB) (Carlesimo, Caltagirone, & Gainotti, 1996), a well standardized battery which explores five cognitive domains through eight neuropsychological tests: visual-spatial intelligence (Raven Coloured Progressive Matrices); controlled verbal production (Verbal fluency test) and verbal competence (Sentence construction); primary and secondary verbal memory (Rey auditory 15-word learning test); simple (copying of drawings: simple copy) and complex (copying of drawings: copy with landmarks) (CDL) constructional function; visual memory (Immediate visual memory). MDB scores range between 0 and 8, and a score inferior to

R. Antonelli Incalzi et al. / Brain and Cognition 70 (2009) 163–170

4 is considered diagnostic of dementia (Carlesimo et al., 1996). MDB was complemented by two tests for immediate memory: visuospatial immediate memory (visual span); short-term memory and attention (verbal span). Raw neuropsychological scores were corrected for age and education according to procedures standardized and validated for the Italian population (Carlesimo et al., 1996; Magni, Binetti, Bianchetti, Rozzini, & Trabucchi, 1996). 2.2. Heart rate variability assessment The study population underwent a 24-h Holter ECG recording by a portable three-channel tape recorder, processed by a Marquette 8000 T system with a sampling frequency of 128 Hz. All recordings were performed at admission: after the preparation of the skin, self-adhesive electrodes were placed in the positions usually used for three-leads Holter monitoring, and recording was started between 9.00 and 9.30 AM. During the recording period the patients were allowed to standing or sitting next to their beds, while other activities were not allowed. In order to be considered eligible for the study, each recording had to have at least 12 h of analyzable intervals between consecutive R peaks. Moreover, the analyzable recording period had to include at least half of the nighttime (from 00:00 AM through to 5:00 AM) and half of the daytime (from 7:30 AM through to 11:30 PM) (Bigger, Fleiss, Rolnitzky, & Steinman, 1992). Each beat was labeled as normal or aberrant according to recognition by the algorithm for tape analysis and after an investigator’s verification. HRV analysis was performed on all consecutive 5 min RR sequences extracted from 24-h holter recordings using a dedicated software (Maestri & Pinna, 1998). From each 5 min sequence HRV indexes were computed, and their mean value across the whole 24 h recording and during daytime and nighttime was obtained (Maestri & Pinna, 1998). Identified RR time series were preprocessed according to the following criteria: (1) RR intervals associated with single ectopic beats were replaced by their mean value, (2) artifacts and runs of tachycardic beats were replaced by N values equal to the mean RR, in such a way that N  mean RR was less than or equal to the substituted value, (3) RR values differing from the preceding one more than 30% (absolute value) were replaced in the same way as for artifacts. The mean RR was computed as a moving average centered on the beat to correct, with a buffer of ±3 beats labeled as normal (Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, 1996). The original and the corrected 5 min sequences were plotted superimposed and the analyst could interactively decide whether to accept or to discard the resulting series. Power spectral density (PSD), a highly reproducible tool to assess the functional balance between parasympathetic and sympathetic domains of the autonomic nervous system activity and to decompose the total variation of a data series into its frequency components, was estimated by the Blackman–Tukey method in all accepted segments after linear trend removal. The total power (TP) and the power in the very low frequency band (VLF, 0.01– 0.04 Hz), low frequency band (LF, 0.04–0.15 Hz) and high frequency band (HF, 0.15–0.45 Hz) were then computed by numerical integration of the spectral density function. The LF/HF ratio was also calculated. Only normalized LF, HF, and LF/HF values were considered in the analysis and expressed as normalized units (n.u.) (Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, 1996). The following HRV parameters were defined in accordance with the ACC/AHA/ESC consensus (Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, 1996): HF component, which reflects mainly the efferent vagal activity; LF component, mediated by efferent vagal

165

and sympathetic activity; LF/HF ratio, a measure of sympatho/vagal balance; VLF component, reflecting neuroendocrine and thermoregulatory influences mediated mainly by the sympathetic system (Takabatake et al., 2001; Vinik & Ziegler, 2007). The most commonly used time-domain parameters were derived from normal RR intervals (NN). Among these, the square root of the mean squared differences of successive RR intervals (RMSSD) is associated with respiratory effects on heart rate and modulated by both parasympathetic and sympathetic activity, the proportion derived by dividing the number of interval differences of successive NN intervals greater than 50 ms (NN50) by the total number of NN intervals (pNN50) reflects rapid adjustments, and the standard deviation of NN (SDNN) express overall HRV regulation (Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, 1996). 2.3. Analytic approach We firstly compared the general characteristics of the subpopulation included in the HRV study with that of excluded patients using Student’s t-test for continuous variables and contingency tables with chi-square for dichotomic ones. Second, we grouped patients according to whether they performed normally or abnormally on the CDL test. Since HRV data were not normally distributed, groups were compared using the Mann–Whitney U test. Third, we grouped patients according to whether they had HRV parameters lower than or greater than/equal to the median value of their distribution, and we used chi-square test to investigate the differences in the prevalence of defective cognitive performances among groups. The relationships between neuropsychological scores and HRV parameters were investigated by Spearman’s rho correlation analysis. Although the basic hypothesis was that drawing impairment correlates with reduced HRV variability, we intended to exclude that such a relationship reflected a generic link between cognitive impairment and depressed HRV. This was the rationale for assessing whether performance in other cognitive domains correlated with HRV. Last, to obtain a further support of the independent correlation between drawing and HRV, we also performed a multivariable linear regression with scores on MMSE and Raven’s Progressive Matrices (two measures of general cognitive performance and general visuospatial intelligence, respectively), as covariates. It is worth mentioning that MMSE has an acceptable diagnostic accuracy (sensitivity 63.2%, specificity 72.2%) in detecting cognitive dysfunction in COPD patients (Antonelli Incalzi et al., 2007), and can thus be considered as an index of general cognitive performance in the present patient population. In order to verify whether autonomic dysfunction was associated with COPD severity, we also used Spearman’s rho correlation analysis to investigate the relationship between HRV parameters and selected indexes of disease severity, such as forced expiratory volume in 1 s (FEV1), PaO2, Barthel Index, Borg’s scale of dyspnea, 6 min walked distance, and number of lost instrumental activities of daily living (IADL). The well proved relationship between deconditioning and impaired HRV was a further rationale to include the 6 min walked distance among potential correlates of HRV (Hasser & Moffitt, 2001). The significance level of correlations was set at p < 0.05. We reasoned that, given the exploratory nature of the study, a higher p level might have concealed true differences more than preventing chance correlations. Moreover, we aimed at verifying a specific working hypothesis (i.e. an association between drawing impairment and reduced HRV), and performed other statistical tests only to exclude a generic link between impaired cognition and reduced HRV. For these reasons, we resolved not to correct p level for multi-

166

R. Antonelli Incalzi et al. / Brain and Cognition 70 (2009) 163–170

ple comparisons. Statistical procedures were performed using SPSS for Win statistical software package V10.0 (SPSS Inc., Chicago, IL, USA).

or VLF power greater or equal to the median value (Fig. 1). No other cognitive test could distinguish LF/HF or VLF based groups (data not shown).

3. Results

Table 1 Demographic and clinical characteristics of patients randomly assigned to the HRV study population and of patients who did not undergo HRV measurement.

Age Gender (M) FEV1 (%) PO2 (mm Hg) Barthel Index, self-care Barthel Index, mobility Barthel Index, total No of lost IADL Borg’s scale of dyspnea 6 min walked distance

No HRV (N = 95)

HRV (N = 54)

p

68.8 ± 9.0 80 (84.2) 36.9 ± 16.5 57.7 ± 8.0 43.8 ± 7.3 39.9 ± 4.7 83.7 ± 10.3 1.6 ± 1.6 5.8 ± 3.4 287 ± 105

69.1 ± 7.7 46 (85.2) 36.0 ± 20.1 56.8 ± 8.8 40.4 ± 8.3 37.8 ± 6.0 78.2 ± 12.8 2.0 ± 1.6 6.9 ± 3.4 253 ± 113

0.880 0.874 0.782 0.529 0.010 0.019 0.005 0.135 0.070 0.079

Data are mean ± SD or number of cases (percentage).

50

40

* 30

20

10

0


median >median 24 hours VLF power, ms2

60

Prevalence of defective performance at the copying drawings with landmarks, %

The patients included in the present HRV study had comparable socio-demographic characteristics but slightly lower Barthel Index scores and 6 min walked distance, and higher burden of dyspnea with respect to patients who did not undergo HRV measurement (Table 1). Visual span, verbal span, and, to a lesser extent, shortterm verbal memory scores were lower in patients included in the HRV study with respect to the excluded ones (Table 2). Thus, both general health and neuropsychological performance were slightly worse in the present patients with respect to the total study sample. A good quality recording of HRV parameters was available for all patients enrolled in the present study. Patients with defective performance at CDL test (N = 23, CDL score range 38.7–61.5) had lower VLF power with respect to patients with normal performance (N = 31, CDL score range 62.1– 70.7), both during the 24 h (median 213; interquartile range 120–282 vs. 309; 188–431 ms2, p = 0.043) and in the daytime period (202; 111–292 vs. 342; 194–397 ms2, p = 0.039), but not in the nighttime period (274; 147–673 vs. 364; 235–689 ms2, p = 0.406). Patients with defective performance also tended to have lower LF n.u. (0.52; 0.44–0.66 vs. 0.62; 0.47–0.74, p = 0.076), higher HF n.u. (0.48; 0.33–0.56 vs. 0.37; 0.26–0.53, p = 0.076), and lower LF n.u./HF n.u. (1.1; 0.8–2.0 vs. 1.7; 0.9–2.8, p = 0.076) during the daytime period. All other time- and frequency-domain HRV parameters did not differ significantly. The prevalence of defective performance at the CDL (score < 61.8) was lower in patients with daytime LF n.u./HF n.u.

Prevalence of defective performance at the copying drawings with landmarks, %

60

50

40

* 30

20

10

0


median >median

24 hours LF n.u./HF n.u. Fig. 1. Prevalence of defective performance at the copying with landmarks test (age- and education-adjusted score < 61.8) in relation to 24 h VLF (upper panel) or LF n.u./HF n.u. (lower panel) values. *p = 0.05 vs.
Table 2 Neuropsychological characteristics of patients randomly assigned to the HRV study population and of patients who did not undergo HRV measurement. Neuropsychological scores

Observed range

No HRV (N = 95)

HRV (N = 54)

p

Raven’s progressive matrices Verbal fluency Visual span Verbal span Verbal memory – short term Verbal memory – long term Copying drawings Copying drawings with landmarks Immediate visual memory Sentence construction No. of defective MDB tests MMSE score

14.3–33.2 9.3–44.5 2–6 2–5 15.0–50.3 1.2–11.8 4.7–13.3 38.7–70.7 10.0–21.3 3.7–30.1 0–6 17–30

24.9 ± 5.4 27.1 ± 8.0 4.5 ± 0.9 4.0 ± 1.4 33.2 ± 8.6 6.2 ± 2.6 10.8 ± 2.0 61.4 ± 8.8 16.7 ± 2.8 18.6 ± 7.5 1.6 ± 1.6 25 ± 3.1

24.0 ± 4.5 25.2 ± 7.9 4.0 ± 0.9 3.5 ± 0.7 30.7 ± 7.8 6.1 ± 2.7 10.5 ± 2.1 61.4 ± 7.3 16.6 ± 2.7 19.0 ± 7.6 1.7 ± 1.6 24.8 ± 3.6

0.297 0.156 0.005 0.016 0.073 0.714 0.366 0.998 0.804 0.726 0.613 0.670

Data in columns 3 and 4 are mean ± SD. MDB, Mental Deterioration Battery. MMSE, Mini Mental State Examination.

167

R. Antonelli Incalzi et al. / Brain and Cognition 70 (2009) 163–170

The correlations between individual neuropsychological scores and selected HRV parameters are reported in Table 3. The CDL score correlated significantly with the VLF power, as well as with the LF n.u, HF n.u., and LF n.u./HF n.u. ratio. The correlations between CDL score and VLF power and LF n.u/HF n.u. are graphically depicted in Fig. 2. No further correlation achieved statistical significance (Table 3), and the time-domain HRV parameters did not correlate with any of the neuropsychologic scores (data not shown). The multivariable linear regression analysis showed that the relationship between VLF power and CDL score was still near significant after adjusting for measures of general cognitive performance and visuospatial intelligence (Table 4). Finally, the study of correlations between HRV parameters and indexes of disease severity showed that LF n.u./HF n.u. also correlated significantly with most of COPD severity parameters (Table 5), while VLF in the nighttime period correlated negatively only with PaO2 (Spearman’s q = 0.31, p = 0.025). 4. Discussion Present data confirm the working hypothesis that neuroautonomic unbalance correlates with an abnormal performance on the CDL test, a neuropsychological test known to be associated with survival in COPD. Furthermore, we found that autonomic control deteriorates as COPD worsens. The fact that PaO2 was a weak correlate of LF n.u./HF n.u. ratio (24 h: p = 0.053; daytime: p = 0.085) does not conflict with this statement because the cumulative exposure to hypoxemia depends upon several factors (e.g. quality of the oxygen therapy, depression of respiratory drive dur-

ing sleep, and effort-related arterial oxygen desaturation) which cannot be summarized by the actual PaO2. Although the small sample size makes this a merely exploratory study, the present results seem worthy of interest because they point at an avenue of research in COPD populations. Indeed, sudden death, which might reflect autonomic unbalance, has been reported to be highly prevalent in COPD (Zulli et al., 2006). In the TORCH study a careful categorization of causes of death identified a 16% prevalence of sudden death (140 out of 875) in a population of 6184 COPD patients followed up for three years (McGarvey, John, Anderson, Zvarich, Wise, 2007). Interestingly, the prevalence of acute myocardial infarction was only 3%, suggesting that arrhythmias related to coronary artery diseases are an unlikely explanation for sudden death complicating COPD (McGarvey et al., 2007). However, there is a distinct lack of information on pathogenesis and clinical correlates of neuroautonomic dysfunction in COPD. The main novel finding of the present paper allowed us to tackle this issue. We found that VLF power, which reflects neuroendocrine influences largely mediated by the sympathetic system (Boneva et al., 2007; Vinik & Ziegler, 2007), was moderately depressed in COPD patients performing abnormally on the CDL test. On the other hand, a trend was evident towards an inverse correlation between CDL score and HF, an indicator of vagal modulation of HRV, as if vagal overcame sympathetic modulation of HRV in CDL impaired patients. Furthermore, VLF power did not correlate with performance on any other neuropsychological test. These observations, although tempered by the findings of the multivariable analysis, converge in supporting the existence of specific VLF-CDL relationships. This

Table 3 Spearman’s correlation coefficients of neuropsychological scores vs. selected HRV parameters. VLF power

LF power n.u.

HF power n.u.

LF n.u./HF n.u.

24 h period Raven’s progressive matrices Verbal fluency Visual span Verbal span Verbal memory – short term Verbal memory – long term Copying drawings Copying drawings with landmarks Immediate visual memory Sentence construction

0.203 0.011 0.090 0.132 0.071 0.036 0.099 0.267* 0.011 0.107

0.041 0.002 0.033 0.118 0.048 0.092 0.178 0.268* 0.018 0.079

0.041 0.002 0.033 0.118 0.048 0.092 0.178 0.268* 0.018 0.079

0.041 0.002 0.033 0.118 0.048 0.092 0.178 0.268* 0.018 0.079

Daytime period Raven’s progressive matrices Verbal fluency Visual span Verbal span Verbal memory – short term Verbal memory – long term Copying drawings Copying drawings with landmarks Immediate visual memory Sentence construction

0.256 0.058 0.132 0.190 0.151 0.038 0.149 0.299* 0.051 0.020

0.108 0.051 0.033 0.143 0.071 0.165 0.205 0.330* 0.053 0.059

0.108 0.051 0.033 0.143 0.071 0.165 0.205 0.330* 0.053 0.059

0.108 0.051 0.033 0.143 0.071 0.165 0.205 0.330* 0.053 0.059

Nighttime period Raven’s progressive matrices Verbal fluency Visual span Verbal span Verbal memory – short term Verbal memory – long term Copying drawings Copying drawings with landmarks Immediate visual memory Sentence construction

0.108 0.095 0.022 0.005 0.065 0.103 0.110 0.078 0.246 0.026

0.021 0.126 0.066 0.160 0.050 0.086 0.134 0.185 0.026 0.124

0.021 0.126 0.066 0.160 0.050 0.086 0.134 0.185 0.026 0.124

0.021 0.126 0.066 0.160 0.050 0.086 0.134 0.185 0.026 0.124

*

p < 0.05.

168

R. Antonelli Incalzi et al. / Brain and Cognition 70 (2009) 163–170

Fig. 2. Relationship between VLF power (A and B) or LF n.u./HF n.u. (C and D) and the score obtained at the copying drawings with landmarks test.

Table 4 Multivariable linear regression analysisa of CDL score to HRV parameters.

Table 5 Relationship between LF n.u./HF n.u. and selected indexes of COPD severity.

B

SE (B)

Beta

t

p

R2

8.933 7.572

4.059 3.883

0.296 0.315

1.757 1.950

0.094 0.057

0.12 0.18

LF n.u./HF n.u. 24 h 0.611 Daytime 0.068

0.378 0.043

0.339 0.388

1.616 1.596

0.114 0.118

0.11 0.09

VLF power 24 h Daytime

a

After adjusting for Raven’s progressive matrices and MMSE scores.

would be in line with a damage of the right insular cortex in the context of a wider COPD-related right hemispheric dysfunction (Royall et al., 2007). Indeed, the right insula damage per se is unlikely to account for CDL dysfunction (Oppenheimer, Gelb, Girvin, & Hachinski, 1992; Trojano & Conson, 2008). Thus, it would be plausible that the diffuse cerebral metabolic impairment well documented in COPD patients (Antonelli Incalzi et al., 2003) might induce, among others, lesions to functional connection of the insular cortex. Indeed, the right insula has reciprocal connections with a great variety of sensory areas (Augustine, 1996), and has a role in the neural network integrating cognitive and emotional aspects of

24 h Spearman’s rho p-Value

FEV1 (%)

PaO2 (mm Hg)

0.321

0.267

0.232

0.294

0.246

0.186

0.406

0.022

0.053

0.095

0.031

0.072

0.188

0.003

0.239

0.167

0.286

0.206

0.170

0.403

0.085

0.232

0.036

0.135

0.229

0.003

0.164

0.288

0.258

0.228

0.037

0.327

0.240

0.037

0.060

0.098

0.799

0.016

Daytime period Spearman’s 0.301 rho p-Value 0.032 Nighttime period Spearman’s 0.272 rho p-Value 0.053

PaCO2 (mm Hg)

Barthel Index – total

Borg’s scale of dyspnea

6 min walked distance (m)

No. of lost IADL

human experience, in order to elicit appropriate responses to stimuli (Wiens, 2005). For instance, right anterior insula, which modulates sympathetic tone, is implicated in the sympathetic arousal associated with mental tasks (Critchley et al., 2000). Thus, the previously observed association between drawing impairment and

R. Antonelli Incalzi et al. / Brain and Cognition 70 (2009) 163–170

mortality (Antonelli Incalzi et al., 2006) might partly reflect depressed responsiveness to stimuli instead of the proarrhythmic effect of neuroautonomic dysfunction. The well recognized primary role of insula in the perception of dyspnoea could also support an alternative hypothesis, according to which both neuroautonomic unbalance and drawing impairment simply behave as markers of insular dysfunction generically reflecting COPD severity (O’Donnell et al., 2007). Accordingly, neuroautonomic dysfunction, foreseen by the present working hypothesis, might not be the only link between right insular damage and death. An important connection between brain and heart has been demonstrated during different cerebral diseases (Samuels, 2007) leading to increased sympathetic discharge and, then, cardiac arrhythmias and even to coagulative myocytolisis (Karch & Billingham, 1986). Similar effects are occasionally seen during severe stress in the absence of organic brain lesions (Cebelin & Hirsch, 1980). On the other hand, severe bradycardia or bradyarrhytmias may complicate conditions rising intracranial pressure (Endo, Sato, Takahashi, & Kato, 2001). All these cardiac abnormalities complicating cerebral problems are expression of the brain-heart connection and might be defined neurogenic heart diseases. Identifying HRV abnormalities in subjects free from clinically evident mechanical and arrhythmic problems might be a clue to detect timely patients at risk of such diseases. This screening might impact clinical management of COPD because arrhythmias are highly prevalent and negatively affect survival (Fuso et al., 1995). If HRV abnormalities will be proved to herald clinically significant arrhythmias, a simple and inexpensive clue to suspect them such as a defective copying test would carry more than a pure neuropsychological information. This study has some important limitations. First, it was designed to assess the effects of cognitive training on a variety of neuropsychological functions, and not to specifically test the interpretive hypotheses for relationships between drawing and mortality. For this reason, we used a broad set of cognitive tests and had only a few measures of executive and praxic abilities. Second, patients who had HRV measured were too few to assess the HRV-survival relationship reliably. Indeed, the correlation between cognitive function and neuroautonomic balance emerged as a secondary objective based on results of the survival study (Antonelli Incalzi et al., 2006). This carries all the conceptual and procedural problems weakening the assessment of correlates of an unplanned outcome. However, the logic intrinsic to present findings points at some biological basis and not at a chance finding. Furthermore, patients who underwent HRV measurement had lower Barthel Index scores and 6 min walked distance with respect to patients not included in the HRV study, which may suggest greater deconditioning and more impaired HRV. Third, the CDL observed range was far more restricted in the normal performance range versus the defective range which might affect our correlation analysis. However, a comparably skewed distribution was present in scores on the short-term verbal memory task and the MMSE, but these tests did not correlate with HRV parameters. This suggests that the CDL–HRV correlation is a true one and not the effect of a statistical bias. Finally, it should be observed that drawing impairment, while highly prevalent, was mild to moderate. It is conceivable that a stronger correlation between HRV abnormalities and drawing impairment would have emerged in a population encompassing the full spectrum of the neuropsychological defect. In conclusion, we observed a correlation between drawing impairment and depressed sympathetic modulation of the neuroautonomic tone. Such a relationship seems worthy of reassessment in a larger and more heterogeneous COPD population in the framework of properly designed studies including also a well balanced set of executive and copying tests. Confirming or denying present findings would make physicians aware of whether a further

169

dimension should be added to the many characterizing COPD as a systemic disease. Acknowledgments The Authors are grateful to Dr. Orsolina Anna Izzo for collection of neuropsychological data and to Dr. Sonia Verna for her precious collaboration. References Allan, L. M., Ballard, C. G., Allen, J., Murray, A., Davidson, A. W., McKeith, I. G., et al. (2007). Autonomic dysfunction in dementia. Journal of Neurology Neurosurgery and Psychiatry, 78, 671–677. Allen, S. C., Jain, M., Ragab, S., & Malik, N. (2003). Acquisition and short-term retention of inhaler techniques require intact executive function in elderly subjects. Age Ageing, 32, 299–302. American Thoracic Society (1995). Standards for the diagnosis and care of patients with chronic obstructive pulmonary disease. American Journal of Respiratory and Critical Care Medicine, 152, s77–s120. Antonelli Incalzi, R., Gemma, A., Marra, C., Muzzolon, R., Capparella, O., & Carbonin, P. (1993). Chronic obstructive pulmonary disease. An original model of cognitive decline. American Review of Respiratory Disease, 148, 418–424. Antonelli Incalzi, R., Gemma, A., Marra, C., Capparella, O., Fuso, L., & Carbonin, P. (1997). Verbal memory impairment in COPD: its mechanisms and clinical relevance. Chest, 112, 1506–1513. Antonelli Incalzi, R., Marra, C., Giordano, A., Calcagni, M. L., Cappa, A., Basso, S., et al. (2003). Cognitive impairment in chronic obstructive pulmonary disease – A neuropsychological and spect study. Journal of Neurology, 250, 325–332. Antonelli Incalzi, R., Corsonello, A., Pedone, C., Trojano, L., Acanfora, D., Spada, A., et al. (2006). Drawing impairment predicts mortality in severe COPD. Chest, 130, 1687–1694. Antonelli Incalzi, R., Corsonello, A., Trojano, L., Pedone, C., Acanfora, D., Spada, A., et al. (2007). Screening of cognitive impairment in chronic obstructive pulmonary disease. Dementia and Geriatric Cognitive Disorders, 23, 264–270. Antonelli Incalzi, R., Corsonello, A., Trojano, L., Pedone, C., Acanfora, D., Spada, A., et al. (2008). Cognitive training is ineffective in hypoxemic COPD: A six-month randomized controlled trial. Rejuvenation Research, 11, 239–250. Augustine, J. (1996). Circuity and functional aspects of the insular lobe in primates including humans. Brain Research and Brain Research Reviews, 22, 229–244. Bigger, T., Fleiss, J., Rolnitzky, L., & Steinman, R. (1992). Stability over time of heart period variability in patients with previous myocardial infarction and ventricular arrhythmias. American Journal of Cardiology, 69, 718–723. Boneva, R. S., Decker, M. J., Maloney, E. M., Lin, J. M., Jones, J. F., Helgason, H. G., et al. (2007). Higher heart rate and reduced heart rate variability persist during sleep in chronic fatigue syndrome: A population-based study. Autonomic Neuroscience, 137, 94–101. Carlesimo, G. A., Caltagirone, C., & Gainotti, G. (1996). The mental deterioration battery: Normative data, diagnostic reliability and qualitative analyses of cognitive impairment. The group for the standardization of the mental deterioration battery. European Neurology, 36, 378–384. Cebelin, M. S., & Hirsch, C. S. (1980). Human stress cardiomyopathy. Myocardial lesions in victims of homicidal assaults without internal injuries. Human Pathology, 11, 123–132. Claus, J. J., Walstra, G. J., Bossuyt, P. M., Teunisse, S., & Van Gool, W. A. (1999). A simple test of copying ability and sex define survival in patients with early Alzheimer’s disease. Psychological Medicine, 29, 485–489. Consensus Conference. (1987). Differential diagnosis of dementing diseases. JAMA, 258, 3411–3416. Critchley, H. D., Elliot, R., Mathias, C. J., & Dolan, R. J. (2000). Neural activity relating to the generation and representation of galvanic skin conductance response: A functional magnetic imaging study. Journal of Neurosciences, 20, 3033–3040. Critchley, H. D., Mathias, C. J., Josephs, O., O’Doherty, J., Zanini, S., Dewar, B. K., et al. (2003). Human cingulate cortex and autonomic control: Converging neuroimaging and clinical evidence. Brain, 126(Pt 10), 2139–2152. Endo, T., Sato, K., Takahashi, T., & Kato, M. (2001). Acute hypotension and bradycardia by medulla oblongata compression in spinal surgery. Journal of Neurosurgery and Anesthesiology, 13, 310–313. Folstein, M. F., Folstein, S. E., & McHigh, P. R. (1975). ‘‘Mini Mental State”: A pratical method for grading the cognitive state of patients for the clinician. Journal of Psychiatry Research, 12, 189–198. Fuso, L., Antonelli Incalzi, R., Pistelli, R., Muzzolon, R., Valente, S., Pagliari, G., et al. (1995). Predicting mortality of patients hospitalized for acutely exacerbated chronic obstructive pulmonary disease. American Journal of Medicine, 98, 272–277. Hasser, E. M., & Moffitt, J. A. (2001). Regulation of sympathetic nervous system function after cardiovascular deconditioning. Annals of New York Academy of Science, 940, 454–468. Karch, S. B., & Billingham, M. E. (1986). Myocardial contraction bands revisited. Human Pathology, 17, 9–13. Kenny, R. A., Kalaria, R., & Ballard, C. (2002). Neurocardiovascular instability in cognitive impairment and dementia. Annals of New York Academy of Science, 977, 183–195.

170

R. Antonelli Incalzi et al. / Brain and Cognition 70 (2009) 163–170

Maestri, R., & Pinna, G. D. (1998). POLYANN: A computer program for poliparametric analysis of cardio-respiratory variability signals. Computer Methods and Programs in Biomedicine, 56, 37–44. Magni, E., Binetti, G., Bianchetti, A., Rozzini, R., & Trabucchi, M. (1996). Mini-mental state examination: A normative study in Italian elderly population. European Journal of Neurology, 3, 1–5. McGarvey, L. P., John, M., Anderson, J. A., Zvarich, M., & Wise, R. A. (2007). TORCH Clinical endpoint committee. ascertainment of cause-specific mortality in COPD: Operations of the TORCH clinical endpoint committee. Thorax, 62, 411–415. McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., & Stadlan, E. M. (1984). Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA work group under the auspices of department of health and human services task force on Alzheimer’s disease. Neurology, 34, 939–944. O’Donnell, D. E., Banzett, R. B., Carrieri-Kohlman, V., Casaburi, R., Davenport, P. W., Gandevia, S. C., et al. (2007). Pathophysiology of dyspnea in chronic obstructive pulmonary disease: A roundtable. Proceedings of the American Thoracic Society, 4, 145–168. Oppenheimer, S. M., Gelb, A., Girvin, J. P., & Hachinski, V. C. (1992). Cardiovascular effects of human insular cortex stimulation. Neurology, 42, 1727–1732. Ortapamuk, H., & Naldoken, S. (2006). Brain perfusion abnormalities in chronic obstructive pulmonary disease: comparison with cognitive impairment. Annals of nuclear Medicine, 19, 529–534. PHS-HCF (1980). International classification of diseases, ninth revision, clinical modifications. Washington, DC: Public Health Service, Health Care Financing Administration. Rosen, W. G., Terry, R. D., Fuld, P. A., Katzman, R., & Peck, A. (1980). Pathological verification of ischemic score in differentiation of dementias. Annals of Neurology, 7, 486–488. Royall, D. R., Gao, J. H., & Kellogg, D. L. Jr., (2006). Insular Alzheimer’s disease pathology as a cause of ‘‘age-related” autonomic dysfunction and mortality in the non-demented elderly. Medical Hypotheses, 67, 747–758. Royall, D. R. (2006). Double jeopardy. Chest, 130, 1636–1638.

Royall, D. R., Chiodo, L. K., Mouton, C., & Polk, M. J. (2007). Cognitive predictors of mortality in elderly retirees: Results from the freedom house study. American Journal of Geriatric Psychiatry, 15, 243–251. Samuels, M. A. (2007). The brain–heart connection. Circulation, 116, 77–84. Takabatake, N., Nakamura, H., Minamihaba, O., Inage, M., Inoue, S., Kagaya, S., et al. (2001). A novel pathophysiologic phenomenon in cachexic patients with chronic obstructive pulmonary disease: The relationship between the circadian rhythm of circulating leptin and the very low-frequency component of heart rate variability. American Journal of Respiratory and Critical Care Medicine, 163, 1314–1319. Task force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology (1996). Heart rate Variability – Standard of measurement, physiological interpretation and clinical use. Circulation 93, 1043–1065. Thayer, J. F., & Brosschot, J. F. (2005). Psychosomatics and psychopathology: Looking up and down from the brain. Psychoneuroendocrinology, 30, 1050–1058. Trojano, L., & Conson, M. (2008). Visuospatial and visuoconstructive deficits. In G. Goldenberg & B. L. Miller (Eds.), Handbook of Clinical Neurology. Third Series (88, pp. 373–392). Amsterdam: Elsevier. Vinik, A. I., & Ziegler, D. (2007). Diabetic cardiovascular autonomic neuropathy. Circulation, 115, 387–397. Volterrani, M., Scalvini, S., Mazzuero, G., Lanfranchi, P., Colombo, R., Clark, A. L., et al. (1994). Decreased heart rate variability in patients with chronic obstructive pulmonary disease. Chest, 106, 1432–1437. Wiens, S. (2005). Interoception in emotional experience. Current Opinion in Neurology, 18, 442–447. Zulli, R., Nicosia, F., Borroni, B., Agosti, C., Prometti, P., Donati, P., et al. (2005). QT dispersion and heart rate variability abnormalities in Alzheimer’s disease and in mild cognitive impairment. Journal of the American Geriatric Society, 53, 2135–2139. Zulli, R., Donati, P., Nicosia, F., De Vecchi, M., Tantucci, C., Romanelli, G., et al. (2006). Increased QT dispersion: A negative prognostic finding in chronic obstructive pulmonary disease. Internal and Emergency Medicine, 1, 279–286.